"I have the uneasy feeling that we are blurring the distinction between
object language, mathematics in this case, and meta language, the
language in which we talk about the object language."
Maybe,
Rich
----- Original Message -----
From: "Bill Wood" <william.wood3 at comcast.net>
To: "Richard Hennessy" <rich.hennessy at verizon.net>
Cc: "Maxima List" <maxima at math.utexas.edu>; "Barton Willis"
<willisb at unk.edu>
Sent: Monday, October 05, 2009 10:55 AM
Subject: Re: [Maxima] Void in Maxima
> On Mon, 2009-10-05 at 10:36 -0400, Richard Hennessy wrote:
>> Let me tone down my statement. There should be a way to have a function
>> return nothing, sort of like programming languages that allow subroutines
>> as
>> well as functions. I don't like having to always return a value, it
>> forces
>> me to return something random that could cause problems later. It is not
>> a
>> panacea but it is a possible improvement in the programming language.
>
> The family of functional programming languages do not have
> "subroutines", only functions, and all functions return values. Both
> the ML and Haskell families have the concept of the "bottom" value,
> which corresponds to "undefined". You can think of "undefined" as a
> type with a single value; it is used to extend partial functions to
> total by having them evaluate to "undefined" for any value not in their
> domain. Functional programmers are quite comfortable with having all
> functions return values.
>
> I have the uneasy feeling that we are blurring the distinction between
> object language, mathematics in this case, and meta language, the
> language in which we talk about the object language. Interestingly
> enough the functional programing language ML started out as the Meta
> Language for an object language of computable functions.
>
> --
> Bill Wood
>